Skip to main content
AI13 March 2026

Mostly coding

28 tasks. March 13, 2026 closed at 98.1x weighted leverage across 855.0 human-equivalent hours in 523 minutes of wall-clock time. Supervisory leverage came in at 438.5x.

That is 21.4 weeks of human-equivalent throughput in 8.7 hours. The ceiling was 576.0x; the floor was 10.0x. 25 of the 28 entries came from a single project.

Task Log

#TaskHuman Est.ClaudeSup.Factor
1AWS Console Simulator: scaffold + engine + 4 service sims + lab runner + starter lab240.0h25m5m576.0x
2Create 50 lab definitions (JSON) and executors (TypeScript) for SCS-C02 and DEA-C0180.0h25m5m192.0x
3Create 44 AWS Console Sim labs (15 CLF-C02 + 29 SAA-C03) - 88 files + registry update120.0h45m5m160.0x
4GCP Console Simulator comprehensive requirements and design document40.0h18m5m133.3x
5Create GCP Certification Labs Master List (245 labs across 11 certs)16.0h8m3m120.0x
6Write 35 free educational domain specs (2286 leaf goals) for AccelaStudy free tier40.0h20m5m120.0x
7Write 5 free domain spec JSON files (Python/JS/SQL/HTML-CSS/Java) with 68-70 leaves each16.0h8m3m120.0x
8Create 60 lab definitions and executors for MLA-C01 MLS-C01 ANS-C0140.0h25m5m96.0x
9Create comprehensive Azure certification labs master list (295 labs across 14 certs)24.0h15m5m96.0x
10Create 87 lab definition and executor files for SAP-C02 and DOP-C02 in avian-console-sim40.0h25m5m96.0x
11Azure Portal Simulator comprehensive requirements and design document16.0h12m3m80.0x
12Create 35 AWS AI lab definitions and executors for AIF-C01 and AIP-C0140.0h35m5m68.6x
13Write 5 free domain specification JSON files (CI/CD, Data Science, ML Concepts, GenAI, Stats/Probability)16.0h18m5m53.3x
14Create 25 DOP-C02 lab definitions and 25 executors for avian-console-sim (50 files total)40.0h45m5m53.3x
15Create 20 AWS SDK client files for avian-console-sim8.0h12m3m40.0x
16Create SAP-C02 labs 16-25 definitions + all 25 executors for avian-console-sim16.0h25m5m38.4x
17Write 5 free domain specification JSON files (Technical Interview/PM/InfoSec/Blockchain/IoT)16.0h25m5m38.4x
18Create 10 SAP-C02 lab definition JSON files matching executor step counts3.0h5m3m36.0x
19UI automation store + TTS + narration bubble + ExecutionContext expansion for AWS Console Simulator3.0h5m3m36.0x
20Write 5 free domain spec JSON files (C# TypeScript Rust Go Swift fundamentals)6.0h12m5m30.0x
21Write 5 free educational domain spec JSON files (DataViz/Patterns/REST/Agile/Testing)8.0h18m5m26.7x
22Rewrite avian-console-sim views: store Record migration + useResources hooks + AWS sidebar nav for VPC/EC2/IAM + CSS modules + visual parity with real AWS Console8.0h22m5m21.8x
23Write 5 free domain spec JSONs (Computer Architecture/Git/CLI/Docker/K8s)8.0h25m5m19.2x
24Fix TS errors in 19 MLA/MLS executor files (updateResource properties + listResources + implicit any)1.5h6m2m15.0x
25Rewrite resource store Map->Record for zustand reactivity + update all consumers2.0h8m3m15.0x
26Write 5 FREE domain specification JSON files (Networking/Linux/Cybersecurity/Cloud/OS)4.0h16m3m15.0x
27Fix TS errors in 13 AIF/AIP/ANS executor files - updateResource signatures and listResources calls1.5h8m3m11.2x
28"Fix 95 TS errors across 38 lab executor/definition files (updateResource signatures, missing imports, undeclared vars)"2.0h12m3m10.0x

Aggregate Statistics

MetricValue
Total tasks28
Total human-equivalent hours855.0
Total Claude minutes523
Total supervisory minutes117
Total tokens4,145,500
Weighted average leverage factor98.1x
Weighted average supervisory leverage factor438.5x
Human-equivalent weeks21.4

Analysis

The highest factor of the day came in at 576.0x and the lowest at 10.0x, a spread of 57.6 times between the two. That is a wide range for a single day, and it usually means the day mixed mechanical work with work that needed real judgement.

The largest single entry accounted for 240.0 of the 855.0 human-equivalent hours, or 28 percent of the day. No single task dominated the total, so the weighted average is representative.

Supervisory time was 117 minutes against 523 minutes of execution, a ratio of about 1 to 4. Supervisory leverage of 438.5x is the figure I find most honest, because it measures the hours I actually spent rather than the hours a machine spent on my behalf.

Every figure here is recorded at the time the work is done rather than reconstructed afterwards. The human estimate is my own judgement and carries the uncertainty that implies; the minutes and tokens are measured. The full dataset, including this day, is available for download.